Slitherine

travling_canuck wrote: ↑Sun May 24, 2020 12:13 pm I'm not sure whether this is possible, or if possible whether it would be considered desirable, to lop off chances below a certain threshold, say 10%?, and reset these chances to 0%?

yes, that is possible and potentially desirable. In fact, currently it lops off the top 10% chance to charge, ie max chance to charge is 90% even if the numbers would put you at or above 100%.

travling_canuck wrote: ↑Sun May 24, 2020 11:48 am Is there historical evidence of massed bowmen:
a) charging other massed bowmen?
b) charging other non-shooting troops?
c) doing either of the above while their ammunition levels are still high?

I'm curious as to how these types of troops actually behaved. If they had lots of arrows left, would they ever charge any enemy, even enemy shooters, or would they just keep shooting? At what point would they toss the bows and charge into combat? And should this depend on their secondary weapon?

By the way, this is another topic for anarchy: shooters not holding their fire, and instead using ammunition on a threatening enemy unit within range, even if their General wants them to preserve their shots for another target, later.

I could try out:
no anarchy charging, or much lower charge chance, for ranged units who still have ammo left

Anarchy shooting is a neat idea, but I'll put it on the backburner for now until we get the rest figured out

Schweetness101 wrote: ↑Sun May 24, 2020 3:35 pm I could try out:
no anarchy charging, or much lower charge chance, for ranged units who still have ammo left

Yes, this is a possibility. A small negative modifier to anarchy charge, I would say. Bit am I right in thinking that if the ranged unit has fired in that turn then there is zero chance for anarchy?

Schweetness101 wrote: ↑Sun May 24, 2020 3:31 pm to help with the discussion in this thread I'll provide the basic numbers in the chance to charge algorithm I've now sent to Pete:

The close charge to chance values between the tiers are probably because the modifiers in the chancetoanarchycharge() function are typically of a greater magnitude than those base anarchy values. Here is a list of what affects chance to anarchy charge and by how much. Perhaps if we hash out what all of these values should be, then we can get a better starting point:

OK, I understand that now.

**Charge chance set to 0 if:
target is foot defending obstacle, or charger is foot defending obstacle, or target is non-fragmented elephants, or charge would disorder charger, or target is in open and charger is medium in rough or difficult terrain, or charger is mounted and target is non fragmented spears or pikes, or charging would put charger under flank threat. Also added if target is broken to have zero charge chance (maybe it should be non zero but for testing purposes and to avoid some potential bugs I have that as zero for now)

Would it make sense to add artillery and baggage trains to this exclusion list?

1) base chance from 'Anarchic' attribute

Currently I have them at 20-15-10-5-0 in the Squads file. The only troop type I have been experimenting with today is Unit 60 Indian archers which I had put down to 2.

2) if less than 50% chance of winning, charge chance goes down 5%, if at significant disadvantage in combat, decrease by an additional 10%, for max -15% chance to charge from combat disadvantage (this is already zeroing out charge chance for bottom three categories)

This one seems very steep to me. I think "high", "medium" and "low" anarchy prone bands should not be zeroed out on this basis. I need to think about the scale though.

3) if greater than 50% chance of winning, charge chance goes up by 5, if at significant advantage in combat, increase by an additional 10%, for max +15% chance to charge from combat advantage

Ditto. Need to think about this as well.

4) if target is ranged and you are not, +10%

OK.

5) if you are ranged, regardless of enemy, -10%

This is already covered in the high-medium-low-very low grading, so I am not sure if we need it again.

6) If at major shooting disadvantage, regardless of whether I am ranged, and I still have some hope of victory in close combat or better, then +20% more likely to charge

Not really understanding this one. Can you give an example of what you mean please?

7) if foot vs mounted shock troops, -40% chance to anarchy charge

Reason for modifier is good.

If mounted and steady and target is fragmented, +10% chance to charge

Yes, good.

9) if superior or above, -10% chance, if raw or below, +10% chance to charge

Very important one.

10) If out of general's command range, +20% chance to charge (so this one for example alone would make low anarchy units more anarchic than high anarchy units within command range with base values of 20,15,10, and 5)

The most important one by far. Obviously we need to think about the scaling.

11) If you are better off receiving a charge than making it, anarchy charge chance goes down 10%, and if much better off receiving than making a charge, it goes down 20%

Not sure we need this one. Anarchy needs to be as simple as we can make it so players can grasp the basic concepts easily for gameplay purposes.

So you can see those values were mostly chosen in the context of base anarchic values of more like 20,40, and 60. Right now they are so large relative to the base values that you don't see much difference between different anarchy tiers of units. Perhaps think about which of the above conditions should be kept, and at what values relative to whatever new base values seem appropriate.

Just so people know, the base values can be changed in about an hour for all units and we can create multiple Squads files until we get the desired combination. If artillery and baggage can be put into the list of charge exemptions at the top then we will need just four categories of indiscipline/anarchy - high-medium-low-very low.

One way to do this would be to say 'ok, a high anarchy unit outside of a general's command range and facing a target it could easily crush should be getting like a 50% chance to anarchy charge, but a low anarchy unit inside command range and on a nice hill where it would rather defend than attack should be getting like a 1%-5% chance to charge. A normal anarchy unit that is receiving fire and is not ranged itself should perhaps be getting a pretty high like 30% chance to anarchy, even if it is otherwise at like 10%, etc...' just making up numbers there, but the point is if you list a number of hypothetical situations like that, we can tweak the numbers to produce those outcomes retroactively, and then test if in practice it works and is how you would like it. Kind of start from the desired end state as it were, rather than inputting different starting values until you get the desired output.

That would be one way of doing it. Another way would be to start of with a basic framework of values for your 4 different categories, say 64-40-20-0 or 100-66-33-0 for high-medium-low-very low and then work through your modifiers individually so you build up the overall picture one thing at a time.

stockwellpete wrote: ↑Sun May 24, 2020 3:50 pm

Schweetness101 wrote: ↑Sun May 24, 2020 3:35 pm I could try out:
no anarchy charging, or much lower charge chance, for ranged units who still have ammo left

Yes, this is a possibility. A small negative modifier to anarchy charge, I would say. Bit am I right in thinking that if the ranged unit has fired in that turn then there is zero chance for anarchy?

yes I think so, because if you've fired on your turn you should have no ap? in which case you cannot charge, and if you cannot make a normal charge you cannot anarchy charge either

stockwellpete wrote: ↑Sun May 24, 2020 4:18 pm ...

if a unit cannot make charges normally, then it cannot anarchy charge either, so artillery and baggage trains should be excluded by default. Even if their base anarchy was set to 100 I don't think they could anarchy charge. Also, if 'Anarchic' attribute is set to 0, they will never anarchy charge with the rules as they are now. So, a separate, no anarchy tier is not really needed, all you want is to ensure their anarchy value is 0 in the squads file. If you want very low but non zero anarchy charge chance, set anarchic to 1 or above.

yes, numbers 2 and 3 are very steep changes for base values of 20-15-10-5-0.

The 15% change for example was 25% of the base value for the old max anarchy of 60. So if we wanted it scaled the same, we would make it a 5% change for 1/4 of a high anarchy of 20, if you follow me. Perhaps give recommendations in scales? Like, any given unit should be 25% more likely than its base value to charge if at advantage. EG, if base value is 12, it would go up to 15, but if base value is 60, it would go up to 75. I would actually rather not do that though. Currently, the % values are going up and down by a number of points. IE, increasing a 20% base chance to charge by 20% makes it 40%, NOT 24% if you follow. We should choose one or the other method of course, and not mix and match those. I like just changing it as 'points' rather than percent of existing value because it is easier to specify the desired percent chance to charge outcome in any given scenario that way. Do you follow what I am talking about here? This is the distinction:

https://sciencing.com/difference-betwee ... 09115.html

We should pick either or of those methods, because it will get confusing very fast if we mix and match them (I am partial to using percentage points as discussed).

the reason for including 5 still is because of shooty lancers, who I thought should have a lower shooting charging chance than otherwise identical lancers, given that you are more likely to want to keep them back to shoot, that they would be more likely to want to stay back to shoot, and that they normally come from more 'civilized' armies like byzantines.

6) is basically if you are receiving a large amount of incoming fire at which you are at significant disadvantage, BUT if you were to charge into melee things would be at least less bad than the shooting disadvantage for you, then go ahead and anarchy charge in. This is to get anarchy charges in rather than just sitting there and tolerating being shot to death turn after turn, which I think is an important and seemingly realistic one to include, but whether the values are correct is debatable certainly. It is distinct from 4, which is just if you are not ranged at all and enemy is ranged at all, regardless of other circumstances or relative combat ability.

11) the code lets you compare the chances of winning if you were to a) initiate the charge (and thus be placed in the square adjacent to the enemy for terrain, height, etc...purposes), vs b) if you were to receive the charge (and thus enemy is moved to square adjacent to you etc...). This lets you see if charging would be much less in your interest than receiving the charge, in which case you are more likely to wait and receive a charge than make one and throw away your advantage (even if you are still at advantage charging, if you are at even greater advantage defending, then stay defending). I could also change it to be like if you have a greater than 50% chance if defending, but a less than 50% chance of winning if attacking, then defend. This is to mitigate the frustrating scenario of a unit anarchically throwing away a nice defensive position (similar to the reason for preventing mediums from charging out of terrain and into the open anarchically). Dunno what the values should be. The idea is also to kind of clandestinely help the player avoid the most frustrating outcomes by preventing them from happening, even if the player doesn't know the exact underlying reason. You could argue either way for this one I guess.

In terms of the retroactive approach, I would like to lay out a few basic scenarios. Let's just start with a high anarchy unit (warriors and warbands etc...) to get a baseline. We'll say these are the odds any given high anarchy unit will charge an enemy each turn that it stands in front of it within charge range without being given an order (assuming the enemy also just stands there). I've just made up some plausible seeming numbers, but remember again, these are for your anarchy roll each turn:

High anarchy unit:
1) Quality/Close Combat advantage, say facing other non light infantry, taking into account all unit POAs but also height advantages etc...:
___a) equal to enemy: 30%
___b) somewhat superior to enemy: 40%
___c) greatly superior to enemy: 50%
___d) somewhat inferior to enemy: 20%
___e) greatly inferior to enemy: 10%
2) If enemy is ranged (and you are not in this case):
___a) if enemy is shooty lancers (mounted shock): 10%
___b) if enemy is massed archer foot (terrible in close combat): 60%
___c) if enemy is skutatoi or assyrian foot (or some decent in close combat ranged foot): 40%
___d) if enemy is non lancer horse archers (ie you can't catch them, but they can still annoy/goad you): 10%
___e) if enemy is light foot (also probably won't catch them): 20%

And how about for a medium anarchy units, like offensive spears and Lancers, imagine a scenario where you would rather defend than attack:
___a) you would beat the enemy attacking them (say going down the hill to meet them on the flats) but would beat them even worse if they attacked up (like superior offensive spears on a hill overlooking enemy normal offensive spears): 10%
___b) you would be about even if attacking, but have a better chance to win than to lose if defending (say offensive spears defending a hill from similar offensive spears): 5%
___b) you would be unable to beat them if attacking, but are at equal or better chance if defending (say cheap spears defending a hill from superior impact foot type of scenario): <5%
___c) because of your POAs you prefer to 'defend' with countercharges, eg Lancers facing nonlight foot: 5% (this would be 0 if those nonlight foot are spears/pikes)

There are of course infinitely many more scenarios, but if we are able to get starting numbers from above as a baseline it would help a lot in modifying everything from there. These are just to give an idea.

**one other thing to think about is that higher quality units (superior and above) are directly less likely to anarchy charge because of the quality considerations, but more likely to anarchy charge indirectly because their combat margins will be better. Does that seem correct?

How about this for simplicity to start with?

Base values for "high", "medium", "low" and "very low" troop types - +60, +40, +20, +0

Out of command start of the turn...……………………………...+40

"Raw" troop type...…………………………………………………………..+30

Target unit is disrupted …………………………………………………..+15

Target unit is fragged……………………………………………………...+30

Veteran/superior troop type…………………………………………..minus 30

Where the tipping point for "anarchy" test is +90


A "high" anarchy unit, like a warband unit, who is out of command radii will always test for anarchy (unless superior or veteran).

Standard unit 60+40 = 100 requires anarchy test
veteran/superior unit 60+40 minus 30 = 70 will not test unless target is fragged (+30=100)

A "medium" anarchy unit, like a lancer cavalry unit, who is out of command radii will only test for anarchy if an extra factor applies, such as “Target is disrupted”

Standard unit 40+40 = 80 so no anarchy test unless +15 for “target is disrupted” 40+40+15=95
Veteran/superior unit cannot reach 90 in any circumstances.

A “low” anarchy unit, like a defensive spearmen unit, who is out of command radii will only test for anarchy if target is fragged

Standard unit 20+40 = 60 so no anarchy test unless target is fragged (+30= 90)

A “very low” anarchy unit can only test if it is raw, is out of command radius, and its target is fragged 30+40+30=100

Note: this last category may not be needed at all and all troop types in it could be given a zero rating.

Once we have this working and tested we could consider adding a few bits and pieces if players were in general agreement about them.

One query - can foot v mounted shock be added to the exclusion list?

I should also say that if it is felt that the tipping point for anarchy being set at +90 is too tight, then it could be relaxed a bit to +75 maybe.

stockwellpete wrote: ↑Sun May 24, 2020 5:29 pm How about this for simplicity to start with?

Base values for "high", "medium", "low" and "very low" troop types - +60, +40, +20, +0

Out of command start of the turn...……………………………...+40

"Raw" troop type...…………………………………………………………..+30

Target unit is disrupted …………………………………………………..+15

Target unit is fragged……………………………………………………...+30

Veteran/superior troop type…………………………………………..minus 30

Where the tipping point for "anarchy" test is +90


A "high" anarchy unit, like a warband unit, who is out of command radii will always test for anarchy (unless superior or veteran).

Standard unit 60+40 = 100 requires anarchy test
veteran/superior unit 60+40 minus 30 = 70 will not test unless target is fragged (+30=100)

A "medium" anarchy unit, like a lancer cavalry unit, who is out of command radii will only test for anarchy if an extra factor applies, such as “Target is disrupted”

Standard unit 40+40 = 80 so no anarchy test unless +15 for “target is disrupted” 40+40+15=95
Veteran/superior unit cannot reach 90 in any circumstances.

A “low” anarchy unit, like a defensive spearmen unit, who is out of command radii will only test for anarchy if target is fragged

Standard unit 20+40 = 60 so no anarchy test unless target is fragged (+30= 90)

A “very low” anarchy unit can only test if it is raw, is out of command radius, and its target is fragged 30+40+30=100

Note: this last category may not be needed at all and all troop types in it could be given a zero rating.

Once we have this working and tested we could consider adding a few bits and pieces if players were in general agreement about them.

One query - can foot v mounted shock be added to the exclusion list?

so, currently, there is not a tipping point for an anarchy test, at which point the random element is introduced with a dice roll or whatever, like you are describing. What you are suggesting here is quite different from what has been implemented, and would involve rewriting most of the chance to charge stuff.

Instead, there is the base anarchy value, + or - situational modifiers, which comes out to a number between 0 and 90 (max 90 regardless of how high), and is treated as the percent chance to charge. Then, that chance to charge is compared with a randomly generated number between 1 and 100, and if the chance to charge is greater than the randomly generated number, the unit anarchy charges.

The way you should think of the base chance to charge, and the modifiers, is as points out of 100. If, for example, the base chance is 20, and the modifiers add another +10, then there is simply a 30% chance to anarchy charge (a 30% chance that 30 is greater than or equal to a randomly chosen number between 1 and 100).

So, in those terms, what should the various values for the posted modifiers be? And, what should the base values be? Given the type of sample scenarios I laid out, what should the chances to charge, after accounting for everything, in any given scenario, on any given turn, be?

Schweetness101 wrote: ↑Sun May 24, 2020 9:20 pm
so, currently, there is not a tipping point for an anarchy test, at which point the random element is introduced with a dice roll or whatever, like you are describing. What you are suggesting here is quite different from what has been implemented, and would involve rewriting most of the chance to charge stuff.

Instead, there is the base anarchy value, + or - situational modifiers, which comes out to a number between 0 and 90 (max 90 regardless of how high), and is treated as the percent chance to charge. Then, that chance to charge is compared with a randomly generated number between 1 and 100, and if the chance to charge is greater than the randomly generated number, the unit anarchy charges.

The way you should think of the base chance to charge, and the modifiers, is as points out of 100. If, for example, the base chance is 20, and the modifiers add another +10, then there is simply a 30% chance to anarchy charge (a 30% chance that 30 is greater than or equal to a randomly chosen number between 1 and 100).

So, in those terms, what should the various values for the posted modifiers be? And, what should the base values be? Given the type of sample scenarios I laid out, what should the chances to charge, after accounting for everything, in any given scenario, on any given turn, be?

OK, I understand this now. I thought there was a threshold for the test and if you went above that then the test would be triggered. I will have a think today about the values we might use for the main categories that I am suggesting in my post.

I think though that we have to keep this as simple as we possibly can so that players are not bamboozled by the detail. The game is complex enough really without us layering on extra dollops of complexity, because in the end that will be self-defeating. If a player was to ask one of us what "anarchy" was and how does it work in the mod, we should be able to answer fairly simply. Something like - it represents indiscipline in an army, very often because command and control has been degraded by losses, that means that some types of units may operate independently of the player. This explanation tells a player that command and control is the key factor and that certain unit types are more prone to "anarchy". I think there are 3 or 4 good situational modifiers that could be added (like being subject to missile fire), but they need to be clear and obvious to the player from the battlefield situation.

Schweetness101 wrote: ↑Sun May 24, 2020 9:20 pm Instead, there is the base anarchy value, + or - situational modifiers, which comes out to a number between 0 and 90 (max 90 regardless of how high), and is treated as the percent chance to charge. Then, that chance to charge is compared with a randomly generated number between 1 and 100, and if the chance to charge is greater than the randomly generated number, the unit anarchy charges.

The way you should think of the base chance to charge, and the modifiers, is as points out of 100. If, for example, the base chance is 20, and the modifiers add another +10, then there is simply a 30% chance to anarchy charge (a 30% chance that 30 is greater than or equal to a randomly chosen number between 1 and 100).

So, in those terms, what should the various values for the posted modifiers be? And, what should the base values be? Given the type of sample scenarios I laid out, what should the chances to charge, after accounting for everything, in any given scenario, on any given turn, be?

Bearing in mind that I want to keep things very simple for v1 "anarchy" rules . . .

We have a 30-20-10-0 scale for the P column in the Squads file representing "high", "medium", "low" and "exempt" categories of anarchy-proneness.

The only modifiers we use at first are . . .
Out of command radius at the start of a turn = modifier +20
Raw troops modifier +20
Superior/Veteran modifier minus 20

Examples
High anarchy units e.g. warband
Basic value in P column = 30

This would mean that a standard warband unit that had not been fully moved by a player at the end of a turn and was within charge range of the enemy would "anarchy" 30% of the time on average. If that unit was "out of command" as well then it would "anarchy" 50% of the time.

The calculation for an out of command, Superior Warband unit that was not fully moved during the players' turn, and remained in charge range of an enemy unit, would be . . .

30 (high anarchy group) + 20 (out of command) minus 20 (superior) = 30% chance to anarchy charge (And if the unit was "in command" then the maths would be 30 minus 20 = 10% chance)

Medium anarchy units e.g. lancer cavalry
Basic value in P column = 20

The calculation for an average lancer, out of command radius, that was not fully moved by the player and remained in charge range of the enemy, would be . . .

20 (medium anarchy group) +20 (out of command) = 40% chance to anarchy. A superior lancer in that situation would only anarchy charge 20% of the time.

Low anarchy units e.g. Roman hastati/principes
Basic value in P column = 10

The calculation for an above average Roman hastati/principes unit, out of command radius, that was not fully moved by the player and remained in charge range of the enemy, would be . . .

10 (low anarchy group) +20 (out of command) = 30% chance to "anarchy" charge. Veteran hastati units in that situation would anarchy just 10% of the time.

I think Very Low anarchy units, which covered missile troops and skirmishers, should be exempt from anarchy in v1.


This is about as complicated as it needs to be in my opinion. A lot of the discussion has been about battlefield situations and would units do this or that, or feel this or that. That might all be superfluous because players have already had their turn to make those sorts of judgements. What we are left with really are those units that did not fully move at the end of a players turn for whatever reason.

As an observer, I like the idea that you should have very simple v1 anarchy rules.

I would, however, add a modifier for being under missile fire. Simply because it’s fun, intuitive, and completely depended on player’s actions. Skirmishing would be a lot more interesting then.

Quivis wrote: ↑Mon May 25, 2020 1:01 pm As an observer, I like the idea that you should have very simple v1 anarchy rules.

I would, however, add a modifier for being under missile fire. Simply because it’s fun, intuitive, and completely depended on player’s actions. Skirmishing would be a lot more interesting then.

"Anarchy" is checked at the end of a player's turn. Only units that have not fired, or have not fully moved may anarchy charge. If a player has not moved a unit that is in missile range then maybe it deserves to be shot at in the next turn as well.

that looks like a good starting point.

This will need some testing to see if it's a problem out of course, but one potential issue I want to bring up is that if anarchy charge chances are set too low, such that they only happen once or twice in a battle, they might start to look less like an anarchic army simulation, and more like randomized unfairness.

I'm not sure if the above values or too high or too low, but, if in testing, we find that we are only getting one or two anarchy charges even in a large battle with anarchy prone units, I think it would end up feeling worse than just not having it. That is, the more anarchy charges that happen, the more the randomness of them is smoothed out across both players and made fair.

Schweetness101 wrote: ↑Mon May 25, 2020 4:44 pm that looks like a good starting point.

This will need some testing to see if it's a problem out of course, but one potential issue I want to bring up is that if anarchy charge chances are set too low, such that they only happen once or twice in a battle, they might start to look less like an anarchic army simulation, and more like randomized unfairness.

I'm not sure if the above values or too high or too low, but, if in testing, we find that we are only getting one or two anarchy charges even in a large battle with anarchy prone units, I think it would end up feeling worse than just not having it. That is, the more anarchy charges that happen, the more the randomness of them is smoothed out across both players and made fair.

Yes, that's fair enough. If, after some testing, we find it is set too low then it is a relatively quick and easy job to adjust the Squads files or to change the modifiers. The really interesting thing for me will be to see how anarchy intersects with reduced command radius and sub-generals only giving command to their starting contingent. We can set up all sorts of tests to generate some stats and invite players to play some matches and give us feedback.

What I found when I was testing yesterday was a wide variation of anarchy charges per turn because of the RNG. Some of the sets were 0/11 or 1/17, but then others with the same units were 4/9 or 3/10. So sometimes the RNG gives clusters of similar results, which in the context of a battle will be very important in terms of the prevalence of anarchy. I think the sweet spot will be where a good player can control even the most anarchy-prone units for most of the time, but will really start to face problems if leaders are killed, particularly those in command of higher anarchy-prone contingents.

Quivis wrote: ↑Mon May 25, 2020 1:01 pm As an observer, I like the idea that you should have very simple v1 anarchy rules.

I would, however, add a modifier for being under missile fire. Simply because it’s fun, intuitive, and completely depended on player’s actions. Skirmishing would be a lot more interesting then.

I agree with this. Not sure if the game "remembers" that a unit was shot at during the prior player's turn, but if so, I think it could be a reasonable boost to the unit's impetuosity the following turn. May not be feasible, though, in the game engine.

And to the argument "if they don't move, they'll be shot at again next turn", I say - exactly. And that's how the soldiers under fire may feel, too. So even if their general, for some reason, wants them to keep "taking it", they themselves may feel otherwise.

travling_canuck wrote: ↑Mon May 25, 2020 10:03 pm
I agree with this. Not sure if the game "remembers" that a unit was shot at during the prior player's turn, but if so, I think it could be a reasonable boost to the unit's impetuosity the following turn. May not be feasible, though, in the game engine.

And to the argument "if they don't move, they'll be shot at again next turn", I say - exactly. And that's how the soldiers under fire may feel, too. So even if their general, for some reason, wants them to keep "taking it", they themselves may feel otherwise.

I think we have to be very careful not to over-complicate "anarchy" in this mod. Often I think that the most elegant solutions are the most simple ones too. Of course, reasonable cases can be made for extra modifiers like "being shot at", or the enemy unit being "fragged" so an anarchy charge is more likely etc - but it is very easy then to start cluttering up the anarchy part of the mod with all sorts of other modifiers for similar reasonable cases. I don't think this sort of cluttering is necessarily helpful in terms of the playability of the mod - and it certainly makes testing the mod more complicated because there will be more variables at work from the outset. Another objection might be that the player had the opportunity to move into contact with the missile unit, or attack the "fragged unit", and didn't choose to do so and therefore "anarchy" shouldn't really bail them out from the consequences of those poor decisions. "Anarchy" has to be primarily a problem for players to solve in the game.

The other thing is that whatever level anarchy is set at in the mod (whether the P column in the Squads file says 30 or 50 for a particular troop type) the key determinant about the amount of indiscipline experienced by an army in a game is the battle plan of the player. For example, using the Gauls v Roman match-up, one player using the Gauls might try to fight a more defensive battle using the large slope/hill at their end of the battlefield even though the Romans have clear skirmisher superiority. The fairly obvious outcome will be that the Roman skirmishers will cause some of the Gallic warbands to charge off the slope where the legionaries will get at them. So "anarchy" will be an important factor in that game. On the other hand, another player using the Gauls will say to themselves, "I have to attack here because my opponent has skirmisher superiority" and form a completely different type of battle plan that keeps the Gallic units moving forward throughout. Anarchy could be very low, or non existent, in this game even though the P column values are exactly the same.